An investigation of the effects of problem-solving on the standardized test proformance of high school students

This study Investigated the effects of problem-solving on the standardized test performance (ACT) of a sample of disadvantaged high school students from Rule High School In northwest, Knoxville, Tennessee. Data were collected on 54 juniors, 23 males and 31 females, who volunteered as subjects for a six week period during the spring of 1984. The sample was randomized into an experimental and control group. One primary and five ancillary research questions were Investigated:

1. Does training in problem-solving significantly affect the ACT subscales (English, mathematics, social studies, and natural science) and composite scores of disadvantaged high school students?

2. Do students who aspire to attend college perform significantly better than those students who aspire to attend vocational school or to enter employment?

3. Does training in problem-solving have a differential Impact on gender?

4. Does training in problem-solving have a differential Impact on racial membership?

5. Does training in problem-solving have a differential Impact on students with high Internal/external locus of control?

6. Does training in problem-solving reduce test anxiety?

The three Instruments employed in this Investigation were the ACT Assessment Sample Test Booklets, 1981 and 1982, the Children's Nowicki Strickland Internal-External Control Scale, and the Test Anxiety Inventory. The independent variables included training in problem-solving, career aspiration, gender, racial membership, locus of control, and test anxiety. The dependent variables were the ACT subscales and composite scores. The data were analyzed by t-tests for independent samples and 2x2 complete factorial analyses of variance (ANOVA) at the .05 level of significance. The results indicated the effects of treatment only to be significant in respect to mathematics. The remaining subscales (English, social studies, and natural science) and composite scores were unaffected. An interactive effect was revealed for treatment and locus of control on the mathematics subscale with internals scoring lower than externals. There were no interactive effects revealed for treatment and aspiration or treatment and anxiety.